Pelvic inflammatory disease (PID) affects over one million American women every year. It frequently results in infertility, ectopic pregnancy, chronic pelvic pain and recurrence. The costs associated with PID and its sequelae have been estimated at over $4 billion per year. Treatment for over three quarters of women with suspected PID consists of antibiotics to be taken on an outpatient basis. However, the effectiveness of outpatient treatment remains untested by comparison to the alternative of inpatient parenteral antibiotic treatment. Outpatient treatment is initially less costly than inpatient, but there has been no consideration of the long-term costs of PID and its sequelae in light of each treatment's effectiveness. We propose to conduct the first clinical trial to directly compare the effectiveness and cost-effectiveness of outpatient and inpatient antimicrobial regimens recommended for the treatment of PID. At five centers, 1200 women with suspected PID will be randomized to inpatient therapy consisting of parenteral cefoxitin and doxycycline for at least 48 hours, followed by doxycycline orally for a total of 14 days, or to an outpatient regimen consisting of a single dose of parenteral cefoxitin and oral doxycycline for 14 days. Once randomized, all recurrences of PID will be treated with the assigned regimen. Baseline data will be collected from all subjects by interview, physical examination, serum testing, cervical culture, and endometrial biopsy. Patients will be followed in-person at days five and 30, and then by telephone every two months for two years post-randomization. The primary comparison of interest between the two treatment groups will be time to fertility and rates of involuntary infertility. We will also compare between treatment groups the secondary short-term outcomes of time to clinical improvement, patient satisfaction with health care, microbiologic cure, elimination of C. trachomatis by polymerase chain reaction and immunofluorescence, and remission of endometritis on histology; and long-term outcomes of frequencies of tubal occlusion in women with involuntary infertility, repeat episodes of PID ectopic pregnancy, chronic pelvic pain, health service use, as well as patients' ratings of their health states. Finally, we will compare in women assigned to the two treatment groups, disease-related direct and indirect costs as a function of the benefits and burdens of each of outcomes. The importance of this research is that a direct comparison of long-term outcomes after inpatient versus outpatient antimicrobial therapy is crucial to any successful effort for preventing the enormously prevalent and morbid reproductive consequences of PID. Additionally, the quantification of costs associated with any reductions in morbidity will allow the generation of rational treatment guidelines.